That said they've done a fairly good job of growing those companies within their area (Aeronet as Enterprise, Linksys as Home) through incremental improvements.

This may be intentional behavior on Cisco's part to control their exposure to large project failures. IIRC there've been cases where Cisco employees couldn't get support to develop something new so they quit and once Cisco saw their success they were acquired back in.

They "outsource" R&D/Innovation to smaller companies and then buy those.

Hence, their special skills as a company are:

1. Identify future growth areas [market segments] and potential leaders.

2. Seamlessly merge the acquired company into Cisco operations.

There was a good article about this in Fortune/Forbes 6-8 years - did not keep the link.

The incentives at large companies basically make certain types of innovations very difficult to create from within. Large companies tend to have a very fixed set of cost structures and margins. That means that products that would be lower margin are almost automatically deprioritized-even with specific CEO attention it's nearly impossible to change resource allocation processes to foster this kind of innovation.

Instead, the recommended option is to create a subsidiary with no existing cost structure or processes. That subsidiary can then focus on making the new product succeed since it's the only way for them to survive.

Once the subsidiary takes off, you usually do not want to fold it back into the main organization-you'll lose all the advantages that allowed it to succeed in the first place! (There are exceptions, and the book goes into detail on when to integrate and when to give the subsidiary autonomy.)

Buying up new companies is another alternative to starting subsidiaries yourself.

largely on the fact that their software was better.

http://newsroom.cisco.com/dlls/2005/corp_011205.html

Airespace wireless for $450MM.

Airespce brain trust cashed out and jumped out and left Cisco's wifi product poor at best...

So many mistakes... they have solidified their offering - but if you were an airespce/cisco customer when that happened -- it was horrific.

http://www.virishi.net/from-happy-hacking-screw-you-story-me...

http://www.dslreports.com/shownews/Meraki-Annoys-Partners-Cu...

Meraki: Buy hardware and a time-based license per unit, typically for three years. No additional costs.

With the Cisco gear you can buy hardware and essentially use it forever just for the upfront cost. With Meraki you buy the hardware then pay on a regular basis to Meraki to continue using it with their cloud control software.

EDIT: just to clarify - very large enterprises like to get their orders directly from Cisco, with a middleman to cover the "non-direct" status in order to apreciate the largest discount they can (typically ~47-49%)... this eliminates the overhead of dealing with ordering out of distribution - but it murders your lead times and deliverability...

Meraki was one of the most up-front straight forward orders: "Yeah it's in stock and in Fremont - we can have it to you next day, or 4 hours, so no need to order a spare MX400..."

Also, their equipment apparently turns into a brick if you stop paying your support contract (because you'd no longer have any way to manage it). This is not really a problem for many customers who have a policy of keeping active support contracts on everything, but it goes against what people are used to.